云南省不同生态水文分区参考作物蒸散量算法适用性评价

Adaptation Evaluation for Reference Evapotranspiration Calculation Methods in Different Eco-hydrological Regionalization of Yunnan Province

将云南省分为3个区域(Ⅰ,滇西—滇西南山原与高山多水区;Ⅱ,滇西北—滇东北山原河谷中水区;Ⅲ,滇中北高原中水-少水区),基于36个气象站点1958—2013年逐日气象资料,以Penman-Monteith法为标准,利用线性回归法、均方根误差、平均偏差和Nash-Sutcliffe系数对Hargreaves-Samani法、Irmark-Allen法、Priestley-Taylor法、Makkink法、1948-Penman法、Penman-Van Bavel法、Turc法、FAO 24 Radiation法和Jensen-Haise法9种算法的计算精度进行对比。结果表明:1948-Penman法在云南省的适用性最强、计算精度最高,FAO 24 Radiation法与JensenHaise法误差较大,其中Ⅰ区适用性最好的是Hargreaves-Samani法,Ⅱ、Ⅲ区1—6月份为1948-Penman法,7—12月份为Priestley-Taylor法;在相对误差空间分布中,Ⅰ区Hargreaves-Samani法、Irmark-Allen法、Priestley-Taylor法、1948-Penman法的相对误差均在20%以下;Ⅱ、Ⅲ区中,Priestley-Taylor法、1948-Penman法的相对误差较小,为0~20%,同时在Ⅲ区中,Irmark-Allen法的相对误差也相对较小;因此,计算云南省的参考作物蒸散量时,整个区域推荐1948-Penman法,滇西—滇西南山原与高山多水区推荐Hargreaves-Samani法,滇西北—滇东北山原河谷中水区和滇中北高原中水-少水区推荐Priestley-Taylor法。

Adaptation evaluation for calculation methods of ETo under the reference evapotranspiration （ET0） conditions of lacking meteorological methods can obtain simplified data and increase the accuracy with the lack of meteorological data in the large area. To obtain the best calculation methods and assess its adaptability, Yunnan Province was divided into three regions, including the mountain plateau and high mountain water abundance region in western and southwestern of Yunnan （ Ⅰ）, the mountain plateau and river valley water middle region in northwestern and northeastern of Yunnan （Ⅱ） and the plateau water middle and shortage region in central and northern of Yunnan （Ⅲ）, and the daily ET0 was calculated in 36 stations from 1958 to 2013. Ten kinds of methods, such as FAO 56 Penman- Monteith, Hargreaves- Samani, Irmark - Allen, Pristley- Taylor, Makkink, 1948 Penman, Penman - Van Bavel, Turc, FAO 24 Radiation and Jensen - Haise, were used to calculate the daily ETo of each station of Yunnan Province. Penman - Monteith method was used as the standard method to calculate the daily ETo because of its accuracy, and the precision of other methods was evaluated by the linear regression coefficient, the daily relative root mean square error, the mean bias error and the coefficient of Nash - Sutcliffe. The results showed that 1948- Penman method had the strongest applicability and the highest calculation accuracy, and Radiation FAO 24 method and Jensen- Haise method had large error. Among all the methods, Hargreaves -Samani method was the best one in region Ⅰ , and 1948 -Penman method was the most appropriate from January to June while Priestley-Taylor method was the most suitable from July to December of each year for regions Ⅱ and Ⅲ. In the spatial distribution of relative error, the relative error of Hargreaves- Samani method, Irmark- Allen method and Priestley- Taylor method was below 20% , so the three methods were appropriate for region Ⅰ, and the relative error of Priestley - Taylor method and 1948 -Penman method was in the range of 0 -20% , as the minimum in regions Ⅱ and Ⅲ, and the relative error of Irmark - Allen method was small in region Ⅱ. When reference crop evapotranspiration was calculated in Yunnan Province, as a consequence, 1948- Penman method was recommended in the whole region, Hargreaves - Saman method was recommended in mountain plateau and high mountain water abundance region in western and southwestern of Yunnan, Priestley - Taylor method was recommended in mountain plateau and river valley water middle region in northwestern and northeastern of Yunnan.